Most of our efforts in linguistic theory are directed towards understanding the nature of productive linguistic processes, and idiosyncrasies are often understood as being a matter for the lexicon. In phonological theory, there is a focus on the characterization of the “phonology proper”, while exceptional forms are taken to be stored in representations that don’t impact upon the phonological grammar of a given language directly; in semantics, the principle of compositionality directs our focus towards compositional processes of maximal generality, with lexical semantic effects being explained in terms of lexical semantic decomposition in syntax, the theorist’s main job is to understand the workings of fundamental operations such as Merge and Agree, and exceptional patterns playing a minor role (if any) in guiding the broader theoretical issues. In these domains, the focus is typically on the ‘core’, with less attention to the ‘periphery’.
Things are a bit different in the domain of morphological theory, where the focus is on listemes and their relationship to other modules; in this area, determining the nature of idiosyncrasies is a core research objective, with major consequences for theory comparison at the framework level. Consider the case of suppletive allomorphy, which is where we see both regular and idiosyncratic forms for what seems to be one and the same syntactic element: the past 15 years has seen a surge in work on the characterization of suppletion, in particular in the wake of Bobaljik (2012)’s groundbreaking work on comparatives and superlatives, and advances in theories of morphology have been driven by this work as the structural conditions on allomorph selection have been refined on the basis of a broadening understanding of the empirical patterns into other domains (Embick 2010; Merchant 2015; Smith et al. 2019; Paparounas 2024; Bešlin 2025; Angelopoulos & Spyropoulos to appear). Relevant for our understanding of these phenomena is the delimitation of different classes of allomorphy, as in some instances we might find that suppletion misnames cases where there is syntactic non-identity (see e.g. Kayne 2018 on ‘se’-based possessive clitics in Romance), and in others there may be more of a role for phonology in determining the range of possible forms than we had realized at first (e.g. Scheer 2016, Newell 2023).
In a similar vein, another branch of research in the Distributed Morphology framework has sought to capture patterns of syncretism, where we see regularity of form but a range of distinct meanings, in terms of allosemy, whereby there is ‘late insertion at LF’ of context-specific denotations (Wood 2012, 2013, 2016; Marantz 2013; Myler 2016; Wood & Marantz 2017; Kastner 2020). These issues have been given a particularly sharp treatment in Wood’s (2023) book on Icelandic nominalizations, in which Wood breathes new life into old questions from Chomsky (1970) about how lexical idiosyncrasies might shape our view of the architecture of the grammar. The allosemy outlook is a natural extension of the treatment of phrasal idioms such as ‘kick the bucket’, and so it is fitting to consider the consequences for this work of recent developments in the study of idioms. Bruening (2010) proposes a theory of idiom formation based on selectional contiguity, and pursuing this work has reinvigorated old debates on the status of functional projections with respect to selection (see e.g. Bruening et al. 2018). Bruening’s work builds on work on ditransitives by Harley (1995, 2002) and Richards (2001), in which recurrence of idiomatic meaning with distinct predicates has been used to argue for lexical decomposition, while the recent response to Bruening’s work in Larson (2017) seeks to reassess the distinction between idioms and collocations, paying special attention to the partial compositionality of some idiomatic expressions. Partial compositionality has also been taken to have some role to play in determining the extent to which a given passive can participate in passivization and relativization (Nunberg et al. 1994, cf. Lebeaux 2009, Folli & Harley 2007, McGinnis 2002), although more recent work that has diversified the discussion with data from beyond English (Wierzba et al. 2023) indicates that the picture may be more complicated than that.
Our understanding of the division between the ‘core’ and the ‘periphery’ has also been reconfigured by Yang’s (2016) work on productivity, in which Yang proposes a calculus, the Tolerance Principle, for determining whether or not a potential rule of their grammar is to become part of the learner’s productive grammar. Yang’s work puts a great deal of stock in evidence from acquisition for what does and does not constitute a truly productive rule, with overregularization (‘gived’, ‘mouses’) as the signature of the acquisition of a productive rule; on this account, and the scarcity of ‘overirregularization’ (overgeneralization from irregular forms, e.g. ‘brung’) in child data tells us that minority rules that the analyst may posit may not, in fact, be true rules of the grammar, with substantial implications for how we assess the boundary between regular rules and `lexicalized’ exceptions. The Tolerance Principle has been applied to a number of domains with some success (Belth et al. 2021; Kodner 2020, 2022; Belth to appear; Thoms et al. 2025), but there is a tension between this work and other results in work which suggest that sublexical generalizations may explain patterns in the phonological selectiveness of certain affixation processes (Gouskova et al. 2015, Gouskova 2025). These competing theories differ fundamentally in how they treat irregularity in morphophonology, and it remains to be seen what empirical phenomena in other domains might be brought to bear on the matter.
The purpose of this special session at NELS 56 is to bring researchers together to discuss these issues and make progress on the broader theoretical issues that they impact upon. Potential topics of discussion include but are not limited to:
- the structural conditioning of suppletive allomorphy
- the status of allosemy and structurally-conditioned idiosyncratic meanings more broadly
- the syntax of idioms and collocations and the line between them
- productivity, minority rules and idiosyncratic forms
- lexical specificity effects in syntax and how to model them
- the syntactic and semantic properties of roots
- the border between compositional and idiosyncratic components of lexical meaning
- overirregularization and sublexical generalization
- idiosyncrasy as a diagnostic for distinct classes (e.g. clitic vs affix)
- psycholinguistic properties of regular vs irregular processes
We welcome work from a range of different research traditions, including psycholinguistics, computational linguistics and acquisition.